pm speciation and sources in mexico during the milagro-2006

Atmos. Chem. Phys., 8, 111–128, 2008www.atmos-chem-phys.net/8/111/2008/© Author(s) 2008. This work is licensedunder a Creative Commons License.

AtmosphericChemistry

and Physics

PM speciation and sources in Mexico during the MILAGRO-2006Campaign

X. Querol1, J. Pey1, M. C. Minguill on1, N. Perez1, A. Alastuey1, M. Viana1, T. Moreno1, R. M. Bernabe2, S. Blanco2,B. Cardenas2, E. Vega3, G. Sosa3, S. Escalona3, H. Ruiz3, and B. Art ınano4

1Instituto de Ciencias de la Tierra “Jaume Almera”, CSIC. C/ Lluıs Sole i Sabarıs s/n, 08028 Barcelona, Spain2Centro Nacional de Investigacion y Capacitacion Ambiental, CENICA. Periferico 5000, Col. Insurgentes Cuicuilco, C.P.04530, Delegacion Coyoacan, Mexico D.F., Mexico3Instituto Mexicano del Petroleo, Eje Central Lazaro Cardenas 152, Col. San Bartolo Atepehuacan. Delegacion Gustavo A.Madero. 07730, Mexico, D.F, Mexico4CIEMAT, Avda. Complutense 22, 28040 Madrid, Spain

Received: 5 July 2007 – Published in Atmos. Chem. Phys. Discuss.: 23 July 2007Revised: 25 September 2007 – Accepted: 5 December 2007 – Published: 14 January 2008

Abstract. Levels of PM10, PM2.5 and PM1 and chemicalspeciation of PM10 and PM2.5 were measured during the MI-LAGRO campaign (1st to 31st March 2006, but extended insome cases until 6th April) at four urban, one suburban, tworural background and two rural sites, with different degree ofindustrial influence, in the Mexico City Metropolitan Area(MCMA) and adjacent regions. PM10 and PM2.5 daily lev-els varied between 50–56µg/m3 and 24–46µg/m3 at the ur-ban sites, 22–35µg/m3 and 13–25µg/m3 at the rural sites,and 75µg/m3 and 31µg/m3 at the industrial hotspot, lowerthan those recorded at some Asian mega-cities and similarto those recorded at other Latin American cities. At the ur-ban sites, hourly PM2.5 and PM1 concentrations showed amarked impact of road traffic emissions (at rush hours), withlevels of coarse PM remaining elevated during daytime. Atthe suburban and rural sites different PM daily patterns wereregistered according to the influence of the pollution plumefrom MCMA, and also of local soil resuspension.

The speciation studies showed that mineral matter ac-counted for 25–27% of bulk PM10 at the urban sites and ahigher proportion (up to 43%) at the suburban and rural sites.This pattern is repeated in PM2.5, with 15% at urban and28% at suburban and rural sites. Carbonaceous compoundsaccounted for a significant proportion at the urban and in-dustrial sites (32–46% in PM10, and 51–55% in PM2.5),markedly reduced at the suburban and rural sites (16–23% inPM10, and 30% in PM2.5). The secondary inorganic aerosolsaccounted for 10–20% of bulk PM10 at urban, suburban, ru-ral and industrial sites, with a higher proportion (40%) at theindustrial background site. A relatively high proportion ofnitrate in rural sites was present in the coarse fraction.

Correspondence to: Xavier Querol([email protected])

Typically anthropogenic elements (As, Cr, Zn, Cu, Pb, Sn,Sb, Ba, among others) showed considerably high levels atthe urban sites; however levels of particulate Hg and crustaltrace elements (Rb, Ti, La, Sc, Ga) were generally higher atthe suburban site.

Principal component analysis identified three major com-mon factors: crustal, regional background and road traffic.Moreover, some specific factors were obtained for each site.

1 Introduction

During the last few decades a large number of epidemio-logical studies have shown a link between pollution by air-borne particulate matter (PM) and respiratory and cardio-vascular disease (Dockery et al., 1993; Kunzli et al., 2000;WHO, 2003). Because a lower threshold under which nohealth effects are observed cannot be determined (WHO,2003 and 2005), toxicological studies are currently aimingto identify which particle characteristics are responsible forwhich adverse health effects (e.g., particle number, size, sur-face, chemical composition). Anthropogenically emitted PMis generally considered to pose the largest threat to humanhealth given its small grain-size, although the effects of nat-urally emitted PM (wind-blown dust, sea salt) cannot be dis-carded (CAFE, 2004).

PM air pollution in urban agglomerations comes mostlyfrom anthropogenic sources (i.e. traffic, industrial processes,energy production, domestic and residential emissions, con-struction), but there is also a minor contribution from naturalsources (e.g. bioaerosols, soil dust, marine aerosol, volcaniceruptions). Once emitted into the atmosphere, this complexmixture of pollutants may be transformed as a function of the

Published by Copernicus Publications on behalf of the European Geosciences Union.

112 X. Querol et al.: PM speciation and sources in Mexico during MILAGRO campaign

Figure 1

99 40oW 99 00oW 98 60oW

19 00oN

19 40oN

19 80oN

20 20oN

T0

T1

T2Jasso

PXT

UNAM

CENICA

Corena

Tenango

Tula

2000

<2000 m

2600 m

3000 m

3400 m

4000 m

>4400 m

N

50 km

Gulf of

Mexico

México City

Gulf of

Mexico

México City

99 40oW 99 00oW 98 60oW

19 00oN

19 40oN

19 80oN

20 20oN

T0

T1

T2Jasso

PXT

UNAM

CENICA

Corena

Tenango

Tula

2000

<2000 m

2600 m

3000 m

3400 m

4000 m

>4400 m

<2000 m

2600 m

3000 m

3400 m

4000 m

>4400 m

N

50 km

Gulf of

Mexico

México City

Gulf of

Mexico

México City

Fig. 1. Location of the PM monitoring sites in and around MexicoCity and the Tula petrochemical estate.

ambient conditions and the interaction among different PMcomponents, as well as gaseous pollutants. The PM systemis especially complex in mega-cities due to:

- Large emission volumes of PM components and gaseousprecursors.

- High variability of sources.- Widespread distribution of emission sources.- Possible long-range transport of the polluted air masses.The WHO/UNEP report on air quality in 1992 focused on

20 mega-cities of the world, which were defined as urbanagglomerations with populations of 10 million or more bythe year 2000 (WHO/UNEP, 1992). The list of mega-citiesincluded not only Tokyo, Mexico City, Cairo, Bangkok orBeijing, but also Los Angeles, New York and London. Mon-itoring air quality in large urban agglomerations is a press-ing need in order to ensure the health and well being of ur-ban residents, but it is also essential if we intend to preventair pollution-related problems from occurring in emergingmega-cities which may influence both air quality and climatechange on the regional, continental and global scales (Molinaet al., 2007). Preventing pollution problems before they oc-cur is usually the most cost-effective method for dealing withair pollution.

In order to better understand the evolution of trace gasesand particulates originating from anthropogenic emissions inMexico City and their impact on regional air quality and

climate, a field campaign called the Megacities Initiative:Local And Global Research Observations (MILAGRO, http://www.eol.ucar.edu/projects/milagro/) collected a wide rangeof meteorological, chemical, gaseous and particulate mea-surements during March 2006. The design of this campaignwas partly based on the results obtained during the MCMA-2003 Campaign (Salcedo et al., 2006; Volkamer et al., 2006;Molina et al., 2007; Zhang et al., 2007), during which ex-ploratory field measurements of ambient aerosols were car-ried out in Mexico City. This campaign helped in definingphysical and chemical properties of the aerosols, as well asthe dominant meteorological scenarios in this mega-city, onwhich the current MILAGRO campaign were based. Dur-ing the MILAGRO campaign measurements were obtainedover a wide range of spatial scales from local, regional, andlarge-scale field experiments to describe the evolution of theMexico City pollutant plume from its source and up to sev-eral hundred kilometers downwind.

The present study deals with the measurements of ambientlevels of TSP, PM10, PM2.5 and PM1, and the speciation andsource apportionment of TSP, PM10 and PM2.5 in the MexicoCity Metropolitan Area (MCMA), by means of an intensivesampling campaign (1st March to 6th April 2006) of mea-surement, sampling and analysis of PM integrated into the2006 MILAGRO campaign. The main goal of the presentstudy is to interpret the variability of PM levels and compo-sition (including trace metals) in Mexico City, as well as toidentify the major emission sources (of PM and trace metals)in the area.

Other studies on PM characterization (some of them withspeciation and source identification) done in Mexico cityare those published by Miranda et al. (1996 and 2005) andAldape et al. (2005), Johnson et al. (2006), Salcedo etal. (2006), Stone et al. (2007), Molina et al. (2007), Mof-fett et al. (2007), Yokelson et al. (2007). Furthermore, Mug-ica et al. (2002) reported on levels of trace elements in PM;and Vega et al. (2001) on the chemical characterization ofre-suspension source emission profiles.

2 Methodology

2.1 Measurements and sampling

Measurement and sampling of PM were carried out at thefollowing monitoring stations (Table 1 and Fig. 1):

T0.

This monitoring station is located to the northwestern part ofthe basin of Mexico City. It is an urban background site influ-enced by road traffic fresh emissions (300 m from four ma-jor roads surrounding it), domestic and residential emissions,but also potentially influenced by local industrial emissionsand from the Tula industrial area (around 60 km to the north-northwest, in the Hidalgo State). This station was equipped

Atmos. Chem. Phys., 8, 111–128, 2008 www.atmos-chem-phys.net/8/111/2008/

http://www.eol.ucar.edu/projects/milagro/

http://www.eol.ucar.edu/projects/milagro/

X. Querol et al.: PM speciation and sources in Mexico during MILAGRO campaign 113

Table 1. Location of the monitoring stations.

Site Coordinates Type of site

UNAM 19◦19′31′′ N 99◦10′51′′ W 2220 m a.s.l. urban backgroundCENICA 19◦21′32′′ N 99◦04′25′′ W 2232 m a.s.l. urban backgroundT0 19◦29′22′′ N 99◦08′22′′ W 2243 m a.s.l. urban backgroundT1 19◦42′14′′ N 98◦57′45′′ W 2270 m a.s.l. suburbanT2 20◦01′14′′ N 98◦54′09′′ W 2542 m a.s.l. rural backgroundTenango 19◦09′18′′ N 98◦51′50′′ W 2377 m a.s.l. rural backgroundCorena 19◦15′52′′ N 99◦18′07′′ W 2300 m a.s.l. urban backgroundJasso 20◦01′02′′ N 99◦18′55′′ W 2125 m a.s.l. rural-industrial hotspotPXT 19◦51′44′′ N 99◦17′43′′ W 2329 m a.s.l. rural close to industry

with high volume samplers fitted with cut off inlets for PM10(Wedding, 68 m3/h) and PM2.5 (MCV, 30 m3/h) from 1stMarch to 4th April 2006 (n=35, 12 h integrated samples), and6th to 29th March 2006 (n=24, 12 h integrated samples), re-spectively. This monitoring site was also equipped with alaser spectrometer (GRIMM 1107, US1) for real time mea-surement of hourly levels of PM10, PM2.5 and PM1, from19th to 30th March 2006.

CENICA.

This is an urban background station, with similar character-istics to T0, but located in the southeastern side of the city ofMexico. This station was equipped with a high volume sam-pler fitted with a cut off inlet for PM10 (Wedding, 68 m3/h)from 1st March to 6th April 2006 (n=38, 12 and 24 h inte-grated samples). This monitoring site was also equipped witha laser spectrometer (GRIMM 1108) for real time measure-ment of hourly levels of TSP, PM10, PM2.5 and PM1, from8th to 29th March 2006.

UNAM.

This is an urban background station, influenced by road traf-fic and residential emissions, with similar characteristics toT0 and CENICA, but located on the southern side of the cityof Mexico. This station was equipped with a high volumesampler and a cut off inlet for PM10 (Wedding, 68 m3/h) from1st to 15th March 2006 (n=7, only 24 h samples).

Jasso (Subestacion) and PXT.

These two rural background sites were located in the vicinityof a heavy industrialized area, the Tula-Vito-Apasco indus-trial corridor, 60 km north-northwest from the Mexico Citydowntown. Jasso and PXT were 6 km southwest and 25 kmsouth from a large refinery and a power plant (Tula estate),respectively. At the two sites PM10 and PM2.5 samples werecollected by using portable low volume samplers (MiniVol)operating at a flow rate of 0.3 m3/h and previously calibratedunder standard conditions. Samples were collected from

00:00 to 24:00 h each day from 24th of March to 20th April2006, with a total number of samples of n=55 for PM10, andn=55 for PM2.5. Additionally, medium volume sequentialfilter samplers (SFS) operated at a flow rate of 6.8 m3/h andequipped with PM2.5 inlets were used to collect 12-hour sam-ples from 7th to 20th April 2006 (n=56).

T1.

This is a suburban background site located around 50 km tothe north of Mexico City, in an area isolated from major ur-ban agglomerations but close to small populated agglomer-ations, and around 500 m from the closest road. The stationwas equipped with three high volume samplers (Wedding,68 m3/h) with TSP, PM10 and PM2.5 cutoff inlets from 1st to29th March 2006 (n=25, 24 and 23, 12 h integrated samples,respectively).

T2.

This is a rural background site located around 90 km to thenorth of the city of Mexico, in the surroundings of a farmisolated from major urban agglomerations, and around 2 kmfrom the closest road. The monitoring site was equipped witha Wedding high volume (68 m3/h) sampler with a PM10 cutoff inlet from 9th to 17th March 2006 (n=8).

CORENA and TENANGO.

At these urban background and rural background sites, re-spectively, two laser spectrometers (GRIMM 1107) for realtime measurements of hourly levels of PM10, PM2.5 and PM1were installed from 1st to 4th April 2006 at CORENA and28th March to 7th April 2006 at TENANGO.

The main objective of the measurements performed at theurban background sites was the characterization and sourceapportionment of urban aerosols in MCMA. Elucidation ofthe origin of particulate mercury in PM10 was of particularinterest here. In contrast, the main objective of data col-lection at the rural background site was to investigate the

www.atmos-chem-phys.net/8/111/2008/ Atmos. Chem. Phys., 8, 111–128, 2008


Figure 2.

99 40oW 99 00 oW 98 60oW

oN

19 40 oN

oN

20 20 oNN

o o o

19 00 oN

oN

19 80 oN

oNN

o o o

oN

oN

oN

oNN

o o o

oN

o

oN

NNNNN

T1

T2

PXT

UNAM CENICA

CorenaTenango

N

T0

NTula

NNNNNNJasso

CENICA

CorenaTenango

N

T0

NNNNNNN

T1

PXT

CENICA

CorenaTenango

N

T0

NTula

NNNNNNJasso

TSP PM10

PM1PM2.5100 µg/m3 100 µg/m3

100 µg/m3100 µg/m3

Gravimetric values Real time values

T1

CENICA

T1

Tula

Fig. 2. Graphical view of the PM (TSP, PM10, PM2.5 and PM1)concentrations shown in Table 2 superimposed onto Fig. 1 (locationmap) to allow easier comparison. Real time values (squares) areonly used when gravimetric values (circles) were not available.

influence of the emissions from MCMA on the compositionof the atmospheric aerosols in the surrounding rural sites.

Samples from T0, T1, T2, UNAM and CENICA obtainedwith the high volume samplers were collected at all siteson quartz micro-fiber filters (Pallflex for the Wedding in-struments and Schleicher & Schuell for the MCV instru-ments), conditioned for subsequent analysis. High volumesamplers were operated one out of every two days along thefield campaign. Two 12 h samples were collected every sam-pling day (from 08:00 h to 20:00 h and 20:00 h to 08:00 h lo-cal time), with the exception of UNAM and T2 (24 h sam-ples). Samples from Jasso and PXT sites were collected on47 mm Teflon-membrane (2µm pore size, Gelman Scientific,Ann Arbor, MI, USA) and quartz-fiber (Pallflex ProductsCorp., Putnam, CT, USA) filters. Teflon filters were usedfor mass analyses, trace element analyses and light transmis-sion, whereas quartz filters were used for ion, elemental andorganic carbon analyses. Although two different types ofquartz filters were used, their quality was checked by eachof the research groups in their respective laboratory.

Real time measurements of PM were obtained with twomodels of laser spectrometers, GRIMM 1107 and 1108.These instruments perform particulate size measurements by

y = 0.8x

R2 = 0.9

y = 1.0x

R2 = 0.6

y = 1.0x

R2 = 0.9

20

40

60

80

100

20 40 60 80 100

T0 PM10 µg/m3

CE

NIC

A P

M1

0 µ

g/m

3

08:00-20:00 20:00-08:00 24 h mean

Figure 4. Fig. 3. Comparison of 12 and 24 h PM10 levels simultaneouslymeasured at T0 and CENICA (two urban background sites), MexicoCity, during the MILAGRO campaign.

90-degree laser light scattering. Sampling air (1.2 l/min)passes through a flat laser beam produced by an ultra lowmaintenance laser diode. A 15-channel (0.3–25µm) pulseheight analyzer for size classification detects the scatteringsignals. These counts from each precisely sized pulse chan-nel are converted to mass using a density equation included inthe instrument software. In addition, real time measurementsobtained by the optical counter were corrected with the re-gression equations (in all cases R2>0.8) obtained betweenthe comparison of gravimetric PM levels (aerodynamic di-ameter) determined by the standard high volume gravimetricmethod, and real time (optical counts) measurements, inde-pendently for PM10 and PM2.5 fractions. PM1 data was cor-rected according to the PM2.5. We have checked at laboratorythat in some cases this may result in a slight overestimationof levels of PM1.

2.2 Chemical and microscopy analysis

Samples were conditioned after sampling (50±4% humid-ity and 22±3◦C during 48 h) prior to gravimetric determi-nation of the PM mass were stabilized (carried out in thesame controlled chamber as filters). Levels of major andtrace elements were determined in bulk sample acidic diges-tions (HF:HClO4:HNO3) by means of Inductively CoupledAtomic Emission Spectroscopy, ICP-AES (IRIS AdvantageTJA Solutions, THERMO), and Inductively Coupled PlasmaMass Spectroscopy, ICP-MS (X Series II, THERMO), re-spectively. Levels of particulate mercury were determinedby means of a Hg Gold Amalgam Atomic Absorption ana-lyzer (AMA-254, LECO instruments). The content of Cl−,SO2−

4 and NO−

3 was determined by means of Ion Chromatog-raphy HPLC (High Pressure Liquid Chromatography) usinga WATERS IC-pakT M anion column and WATERS 432 con-ductivity detector, and NH+4 was determined by means ofa Selective Electrode (MODEL 710 A+, THERMO Orion).Total Carbon was determined by means of a LECO ele-mental analyzer. OC and EC levels were determined on aselected number of samples to determine the ratios OC/EC



0

25

50

75

100

0 2 4 6 8 10 12 14 16 18 20 22

local time (hour)

PM

(µ

g/m

3)

PM10 PM2.5 PM1

CENICA

PM10 PM2.5 PM1

0

25

50

75

100

0 2 4 6 8 10 12 14 16 18 20 22

local time (hour)

PM

(µ

g/m

3)

PM10 PM2.5 PM1

T0

PM10 PM2.5 PM1

0

25

50

75

100

0 2 4 6 8 10 12 14 16 18 20 22

local time (hour)

PM

(µ

g/m

3)

PM10 PM2.5 PM1

CORENA

PM10 PM2.5 PM1

0

25

50

75

100

0 2 4 6 8 10 12 14 16 18 20 22

local time (hour)

PM

(µ

g/m

3)

PM10 PM2.5 PM1

TENANGO

PM10 PM2.5 PM1

Fig. 4. Mean hourly PM levels measured during the MILAGRO campaign at 2 urban background sites of Mexico City, CENICA and T0,and at two southern sites, Tenango and Corena.

by Thermo-Optical Reflectance (TOR) according to themethod described by Fung et al. (2002). Sample treatmentand analytical details are described in Querol et al. (2001).Indirect determinations from analytical data were obtainedfor: a) CO2−

3 , determined from Ca content, assuming thatthis element is mainly present as calcite (CaCO3; CO2−

3 =1.5*Ca); b) SiO2, determined from the Al content on thebasis of prior studies in Mexico city (SiO2=2*Al 2O3; John-son et al., 2006); c) organic matter, determined as 1.6*OC,based on Turpin and Lim, 2001). Selected samples were alsostudied under an environmental Scanning Electron Micro-scope (SEM, FEI QUANTA 200), with chemical analyses ofindividual particles being performed manually on uncoatedsamples using an energy dispersive X-ray microanalysis sys-tem (EDX). Microscope conditions were working distance of10 mm, accelerating voltage of 20 kV, a beam spot size num-ber 2 with a beam current of approximately 1.00µA, and aspectrum acquisition time of 30 s live time, with particles be-ing analyzed in its centre. The time under vacuum (130 Pa)before the analysis is about 2–3 min.

2.3 Meteorological data

At T0 site, a 3-D sonic anemometer (Campbell Scientific,model CSAT3) was used to measure horizontal and verti-cal wind components at 1 Hz. Temperature, relative humid-ity and atmospheric pressure were registered using VaisalaHMP-series sensors. All equipments were deployed at 12 mabove the ground level. Data was kindly made available forMILAGRO participants by W. Eichinger from the Universityof Iowa.

Figure 5

0

90

180

270

360

0 2 4 6 8 10 12 14 16 18 20 22

local time (hour)

Win

d d

irecti

on

(º)

0

5

10

15

20

25

30

35

10 *

Win

d s

peed

(m

/s)

T (

ºC)

Wind direction 10*Wind speed T

Fig. 5. Mean hourly temperature, wind direction (vectorial mean)and wind speed recorded at T0 from 1st to 26th March 2006.

3 Atmospheric dynamics

Mexico City is located in an elevated basin, 2240 m abovemean sea level (m a.s.l.). Surrounding mountains on the west,south and east reach 1000 m above the basin, and two volca-noes at the southeast reach over 5000 m a.s.l. The basin isopened to the north and southeast, where a small gap be-tween mountains allows entering of cold and fresh air duringevenings, cleaning up the basin most of the time. Fast andZhong (1998) emphasize the importance of vertical mixingand mountain winds in the transport of the Mexico City ur-ban plume first towards the south during the day and thenback over the city to the north during nightime. Clear skies,low humidity, and weak winds aloft associated with high-pressure systems are usually observed over Mexico duringMarch (Fast et al., 2007).



Table 2. Mean TSP, PM10, PM2.5 and PM1 (µg/m3) levels obtained at the different monitoring stations, for the periods indicated.

Location n (TSP, PM10, PM2.5) TSP PM10 PM2.5 PM1Real time measurements

CENICA 08–29 March 2006 144 50 24 19T0 19–30 March 2006 56 40 33Tenango 28 March 2006–06 April 2006 22 13 10Corena 01–04 April 2006 30 25 17

PM samplers (gravimetry)CENICA 01 March 2006 to 06 April 2006 38 53T0 01 March 2006 to 06 April 2006 37, 27 50 40T1 01–29 March 2006 23, 24, 25 193 80 33UNAM 01–15 March 2006 7 54T2 09–17 March 2006 8 34Tula- PXT 24 March 2006–20 April 2006 28, 42 37 26Tula- Jasso 24 March 2006–20 April 2006 28, 42 75 31

Based on MILAGRO field measurements and large-scaleanalyses Fast et al. (2007) defined three regimes that char-acterized the overall meteorological conditions: before 14thMarch (mostly sunny and dry conditions), between 14th and23rd March (increase in humidity and development of late af-ternoon convection), and after 23rd March (higher humidity,afternoon convection, and precipitation). The synoptic con-ditions during March for Mexico City are described in de-tail by Fast et al. (2007), with winds from the north and eastduring the first week, west then southwest from 9–12th, vari-able from 13–18th, southwest between 19–20th, and finallywesterly winds for the rest of the month. Wind speeds at500 hPa exceeded 15 m/s during 9th–11th March and 19th–20th March, whereas at 700 hPa they were usually less than5 m/s.

According to Doran et al. (2007), on 10th, 18th, 19th,20th, 22nd, and 24th March T1 and T2 were influenced bythe Mexico’s emission plume. This type of air mass circula-tion coincides with that reported by Molina et al. (2007) dur-ing episodes of ozone transport towards the northeast fromMCMA.

4 Results

4.1 PM levels

Table 2 summarizes the mean TSP, PM10, PM2.5 and PM1levels measured at the different monitoring stations (also seeFig. 2). Although the measuring periods were not simulta-neous at all the monitoring sites, it is evident that the meanPM10 levels measured at the urban background sites of Mex-ico City for the study period fall in a narrow range, from 50to 56µg/m3, the values from real time and gravimetry be-ing very similar. For PM2.5 and PM1 a wider range of levelswas measured, with 24 to 40 and 19 to 33µg/m3, respec-tively. At the Jasso rural-industrial site PM10 levels measuredincreased up to 75µg/m3, although PM2.5 levels (31µg/m3)

were in the range defined for Mexico City (35µg/m3) (Vegaet al., 2004). At the rural sites PXT and T2 PM levels fallwithin the range 34 to 37µg PM10/m3 and 26µg PM2.5/m3.Levels at the suburban site T1 (193, 80 and 33µg/m3 forTSP, PM10 and PM2.5) and the Jasso rural-industrial site weremarkedly higher probably due to intensive local soil dust re-suspension and to the influence of a cement plant and lime-stone quarry at Jasso. It should be noted that at the urbansites, the instrumentation was located on the terrace of build-ings and that coarse PM concentrations (e.g. TSP) at groundlevel would probably be much higher.

Although the above PM levels may be considered as rel-atively high when compared with the US or European airquality standards, they are markedly lower than PM lev-els measured at Asian mega-cities (or large cities). Thus,as shown in Table 3, most large cities in China and Indiarecord annual mean levels of 80–150µg PM10/m3 and 60–125µg PM2.5/m3. Moreover, data reported by Clean Air Ini-tiatives for Asian cities (2006) show that, with the excep-tion of Tokyo, most Asian mega-cities (>10 million popula-tion) and large cities exceed 80µg PM10/m3 as a mean an-nual average (mean 2005 annual levels inµg PM10/m3: 51Bangkok; 155 Beijing; 115 Hanoi, 80 Jakarta; 110 Calcutta;80 Mumbai; 160 New Delhi; 70 Seoul; 100 Shanghai; 35Tokyo). These levels are up to 3 times higher than thoserecorded in Mexico City during the MILAGRO campaign,although it must be taken into account that the PM concen-trations presented in this study are not annual means, as it isthe case for the Asian cities described above.

Similar levels of PM10 were measured at the two urbanbackground sites with the largest data coverage (CENICAand T0), with a simultaneous variation in the levels of PMmeasured at both sites (α=1.0, R2=0.9: see Fig. 3). Daylight12 h PM10 levels were similar (α=1.0, R2=0.6). The vari-ation of PM10 levels was even better correlated at the twosites during the night (R2=0.9), although 20% higher values(α=0.8) were recorded at T0. The higher night-time values



at T0 could be related to aged pollutant transport towards thenorth, as described by Fast and Zhong (1998). These resultsindicate that both stations are indeed representative of urbanbackground PM10 climate in Mexico City and that, in agree-ment with results from Molina et al. (2007) for MCMA-2003campaign, PM10 levels varied on a relatively large scale dur-ing the MILAGRO campaign. The lower diurnal correlationof the PM10 levels is probably attributable to the influenceof soil resuspension at CENICA. In this context, an impor-tant difference between the relative contributions of coarseand fine PM was detected. PM10 at CENICA had a largerproportion of coarse PM (50% PM2.5−10) whereas at T0 thefiner fraction was dominant (only 30% PM2.5−10), probablydue to higher traffic exhaust emissions in the vicinity of T0.

At the suburban and rural sites the coarse fraction ac-counted from 30 to 60% of the PM10 mass. This large vari-ation in the relative proportion of fine and coarse fraction isdue to a high impact of local soil resuspension at T1 and toresuspension of dust from a nearby cement plant and lime-stone quarry at Jasso. At T1 and Jasso sites, these large resus-pended dust contributions accounted for high levels of PM10and PM2.5, when compared with the other rural sites (by afactor of 2 to 4). The fraction PM1−2.5 was about 20% oftotal PM2.5 in the urban sites, which is similar to the findingsreported by Salcedo et al. (2006).

The hourly evolution of PM levels measured during theentire MILAGRO campaign at the urban background sites ofMexico City, CENICA and T0 (Fig. 4), was most stronglyinfluenced by the peak fine PM concentrations recorded dur-ing the morning traffic rush hours. This pattern coincideswith data from Chow et al. (2002) as well as the report of themonitoring network in Mexico City (RAMA) during April2004 at the Metropolitan University as reported by Salcedoet al. (2006). The coarse fraction increased also at morn-ing rush hours, remaining high and reaching the maximumat 16–18 h, followed by a nightly decrease (less urban dustresuspension). The 2 monitoring sites followed the samemean daily pattern, with a very clear afternoon coarse PMpeak, probably attributable to the higher wind speed typicallyrecorded at Mexico City during the afternoon in this periodof the year (Fig. 5).

At the two sites located to the South of Mexico City, differ-ent PM daily trends were recorded, with the maximum PMbeing registered in the afternoon, at 19:00 h at Tenango andat 14:00 h at Corena. This is probably due to the transport ofthe urban plume from Mexico, which reaches Corena (south)at 14:00 h and Tenango (southeast) at 18:00–19:00 h. Bothfine and coarse fractions are affected by this transport. Thisis supported by the measurements at the rural site Tenangoplotted in Fig. 6, showing Mexico City’s plume impact caus-ing high levels of PM1 during the period 15:00–21:00 h andmuch reduced over the rest of the day.

As expected, PM10 levels were at all sites higher (by afactor around 1.3) during daylight, as a consequence of the

0

50

100

150

200

250

300

350

400

03/04/06

12:00

04/04/06

00:00

04/04/06

12:00

05/04/06

00:00

05/04/06

12:00

06/04/06

00:00

06/04/06

12:00

07/04/06

00:00

07/04/06

12:00

PM

(µ

g/m

3)

PM10 PM2.5 PM1

TENANGO

PM10 PM2.5 PM1

Fig. 6. Hourly PM levels measured at the rural site Tenango, show-ing the Mexico City’s plume impact causing high levels of PM1.

higher atmospheric emissions and turbulence. This differ-ence was much reduced in the case of PM2.5.

Concerning the daily evolution of PM levels measured atthe two urban background sites, CENICA and T0 (Fig. 7),two definite periods can be clearly identified. The first onelasted from the beginning of the measurements until 23rdMarch, and was characterized by relatively high levels ofPM10 (a mean of 57µg/m3), with a dominant coarse grainsize (PM2.5−10 and PM1−10 loads were 57% and 67% of bulkPM10) due probably to the influence of resuspension of urbandust. The second period started on 23rd March and lasteduntil the end of the month, and was characterized by lowerPM10 levels (35µg/m3) and a finer grain size (PM2.5−10 andPM1−10 loads were 41% and 50% of bulk PM10). This wascoincident with the last meteorological period of the fieldcampaign when precipitation was frequent and the resuspen-sion of urban dust diminished. This wet period also influ-enced PM2.5 and PM1 levels which slightly decreased (by18%), but not as much as PM10 (by 40%). In addition tothe above described mean hourly patterns (defined by roadtraffic emissions and wind velocity/dust resuspension), twoadditional types of PM episodes were identified. One from04 to 06 h with relatively high levels of PM1 occurring on14th, 18th, 21st, 22nd and 25th to 27th March 2006. Duringmost of those days, PM2.5 levels were higher during the nightthan during daytime. These episodes are also reflected in themean hourly evolution obtained for the urban sites (Fig. 4),and as interpreted already by Doran et al. (2007) on the basisof real time measurements of OC and EC at T1 and T2, theymay be attributed to the trapping of pollutants in the Mex-ico City area overnight and during the morning hours in ashallow surface layer before the rapid growth of the mixinglayer commences around 07:00 h. Another single episode ofcoarse PM10 occurred from 10:00 to 20:00 h on 19th March2006 at CENICA and T0, probably caused by local dust re-suspension due to high wind velocity as recorded by Fast etal. (2007) during this day, compared with the rest of the cam-paign.



Table 3. Mean PM10 and PM2.5 levels and PM10 speciation from large Asian cities compared with the data obtained in this study for MexicoCity.* Only water soluble fraction.

Seoula Tokyob Taiwanb China(c−g) Chinah Calcuttai New Delhij,k Mumbail S. de Chilem Buenos Airesn Sao Paulom Mexico This study

µg/m3 min max Nanjing min max CENICA T0PM2.5 56 NR NR 50 127 222 NR NR NR NR 68 35 35 24 46PM10 109 35 NR 79 197 317 140 197 278 NR 115 50 77 52 52OC+EC NR NR NR 14 31 29* 2 3 NR NR NR NR NR 14 16Al2O3 NR NR NR 5 11 NR NR NR NR NR NR NR NR 3 2Ca NR 1 2 2 10 5* NR NR NR NR NR NR NR 1 2Fe NR 1 1 1 3 NR 0.1 0.1 NR NR NR NR NR 1 1K NR 0.4 1 2 4 3* NR NR NR NR NR 1 NR 1 1Mg NR 0.2 1 1 2 1* NR NR NR NR NR NR NR 0.5 0.4Na NR 1 8 1 8 4* NR NR NR NR NR 2 NR 0.5 0.5SO2−

4 NR NR 10 15 30 18 1 2 NR NR NR 7 NR 5 5NO−

3 NR NR NR 3 20 9 0.2 0.2 NR NR NR 4 NR 4 4Cl− NR NR NR 1 5 2 1 1 NR NR NR 5 NR 1 1NH+

4 NR NR NR 4 13 11 NR NR NR NR NR 3 NR 2 2ng/m3

As NR 2 50 6 66 NR NR NR NR 6 NR NR NR 5 6Cr NR 6 20 6 40 NR 7 6 280 100 NR NR NR 1 4Zn NR 233 380 295 1409 NR 490 535 600 NR NR NR NR 100 482Sr NR NR 10 30 40 NR NR NR NR NR NR NR NR 9 16Pb NR 64 90 82 409 NR 40 120 660 900 NR NR NR 32 111Ni NR 5 50 4 75 NR 7 8 250 160 NR NR NR 3 5Co NR 1 – 1 5 NR NR NR NR NR NR NR NR 0.5 1Cd NR 2 10 9 11 NR 2 5 80 NR NR NR NR 1 3Mn NR 30 70 23 186 NR 2 2 NR 860 NR NR NR 20 32V NR 6 10 5 18 NR NR NR NR NR NR NR NR 19 25Cu NR 27 250 40 170 NR NR NR NR 900 NR NR NR 75 110Ti NR 40 NR 80 285 NR NR NR NR NR NR NR NR 114 81

a Kim and Kim (2003) h Wang et al. (2003)b Clean Air Initiatives for Asian cities (2006). Air Quality in Asian cities, 6 p.,i Karar et al. (2007)

http://www.cleanairnet.org/caiasia/1412/article-59689.html j Monkkonen et al. (2004)c Zheng et al. (2004) k Balachandran et al. (2000)d Hea et al. (2001) l Kumar et al. (2001)e Chan et al. (2005) m Rojas-Bracho et al. (2002)f Ho et al. (2003) n Bogo et al. (2003)g Sun et al. (2004) NR Not reported

4.2 PM composition

As shown in Table 4, TSP, PM10 and PM2.5 levelsin Mexico City are highly influenced bycrustal mate-rial (SiO2+CO2−

3 +Al2O3+Ca+Fe+Mg+K): 25–27% of PM10

(13–15µg/m3) in urban sites, and 36 and 43% in the sub-urban site T1 and at the industrial hotspot Jasso (36 and27µg/m3, respectively); confirming that the high PM10 levelsrecorded at these sites are partly due to dust resuspension. InPM2.5 this contribution decreased but still reached relativelyhigh levels (15–28%, 6–9µg/m3 at most sites), this being inagreement with previous works (Miranda et al., 1996; Vegaet al., 2001). This crustal influence is also observed whenstudying the samples under the scanning electron microscope(SEM). Approximately 50% of the particles observed areminerals, mostly silicates (quartz, clays, micas, feldspars),but also phosphates, carbonates (calcite and siderite) and sul-fates (gypsum). Applying the algorithm used by Chow etal. (2002): (1.89xAl+2.14xSi+1.4xCa+1.43xFe) the mineralmatter content for these samples is about 18% less than whencalculating it with our methodology, although the correlationcoefficient between the two methods is very high (R2=0.99).

Thecarbonaceous aerosols (OM+EC) also accounted fora high proportion of the PM mass: 32–46% PM10 (18–23µg/m3, with the highest values for T0) in urban sites, 35–37% (14–27µg/m3) at both industrial influenced sites, andless than 23% (8–13µg/m3) in the suburban and rural sitesT1 and T2. In PM2.5 this contribution increased up to 55%and 30% in the urban and suburban sites, respectively, andfrom 51 to 53% at the industrial sites. The ratio EC/OC inPM10 determined in the present study varied from 0.2 to 0.6for the different sites (with the exception of T1, where theratio was close to 1). An EC/OC ratio of 0.2 for T1 was re-ported by Doran et al. (2007). The average EC/OC ratio atTula was 0.4 (for PM2.5 and PM10), which is consistent withthose values reported by Vega et al. (2007) for Salamanca(0.4 and 0.3 for PM2.5 and PM10, respectively) which is alsoan industrial area influenced by a refinery and a power plant.

At the urban sites of CENICA and T0 (PM10), and at thesuburban site T1 (PM2.5) there is a significant correlationof K and OC+EC (R2=0.4–0.6), but also of K and Al(R2=0.5–0.8) pointing to the relevance of both biomass com-bustion and mineral dust (K-feldspars and illite) emissions


http://www.cleanairnet.org/caiasia/1412/article-59689.html


0

50

100

150

200

250

07/03/06 11/03/06 15/03/06 19/03/06 23/03/06 27/03/06 31/03/06

µg

/m3

PM10 PM2.5 PM1 T0PM10 PM2.5 PM1

0

50

100

150

200

250

07/03/06 11/03/06 15/03/06 19/03/06 23/03/06 27/03/06 31/03/06

µg

/m3

PM10 PM2.5 PM1PM10 PM2.5 PM1

PM10 PM2.5 PM1

CENICA

0

50

100

150

200

250

07/03/06 11/03/06 15/03/06 19/03/06 23/03/06 27/03/06 31/03/06

µg

/m3


0

50

100

150

200

250

07/03/06 11/03/06 15/03/06 19/03/06 23/03/06 27/03/06 31/03/06


PM10 PM2.5 PM1

CENICA

0

50

100

150

200

250

07/03/06 11/03/06 15/03/06 19/03/06 23/03/06 27/03/06 31/03/06

µg

/m3


0

50

100

150

200

250

07/03/06 11/03/06 15/03/06 19/03/06 23/03/06 27/03/06 31/03/06

µg

/m3


PM10 PM2.5 PM1

CENICA

0

50

100

150

200

250

07/03/06 11/03/06 15/03/06 19/03/06 23/03/06 27/03/06 31/03/06

µg

/m3


0

50

100

150

200

250

07/03/06 11/03/06 15/03/06 19/03/06 23/03/06 27/03/06 31/03/06


PM10 PM2.5 PM1

CENICA

Fig. 7. Hourly PM levels measured during the MILAGRO campaign at 2 urban background sites of Mexico City, CENICA and T0.

contributing to the variability of the levels of these com-ponents. The influence of biomass combustion on OC+EClevels also diminishes the typical urban correlation betweennitrate and OC+EC levels. Biomass burning episodes andtheir important influence on PM levels and composition inMexico were previously reported by Johnson et al. (2006),Salcedo et al. (2006), Yokelson et al. (2007), Moffett etal. (2007), Molina et al. (2007), Stone et al. (2007) and Zanget al. (2007). These studies provided quantitative estimationsof the contribution from biomass burning to PM levels, whichvary widely according to the different authors. As describedabove, the influence of biomass burning on PM compositionwas detected in the present study, even though the currentdata does not enable us to provide an accurate estimationof the magnitude of this source contribution. By applyingthe factors relating biomass burning carbon with total OC byStone et al. (2007) to the data from the present study, our re-sults suggest that biomass burning contributions to PM2.5 atT0 should range between 5–15%.

Secondary inorganic aerosols (SIA= SO2−

4 +NO−

3 +NH+

4 )accounted for 15–20% of PM10 and PM2.5 mass at the ur-ban sites, 10–19% at the suburban site, and 20–40% at the

industrial sites, with sulfate accounting for around 50–65%of SIA. Levels of ammonium (1.1–1.5µg/m3 in PM10) andsulfate (4–5µg/m3 in PM10) were homogenous at the differ-ent urban, suburban and rural sites, and relatively higher atthe industrial sites (2.0–2.5µg NH+

4 /m3 and 6.5µg SO2−

4 /m3

in PM10), this low spatial variability agrees with Salcedo etal. (2006), who concluded that sulfate in MCMA has mostlya regional character. Nitrate levels were relatively low, buthigher at the urban sites (2.8–3.6µg/m3 in PM10 and PM2.5),compared with the suburban and rural (1.1-2.8µg/m3) andindustrial (0.6–1.8µg/m3) sites. Levels of SIA in PM2.5 wereonly slightly reduced with respect to PM10 for SO2−

4 andNO−

3 in the urban sites. These results coincide with thoseobtained for PM2.5 at the CENICA site during the MCMA-2003 Campaign, when similar levels of sulfate (3.1µg/m3),nitrate (3.7µg/m3) and ammonium (2.2µg/m3) were regis-tered (Salcedo et al., 2006). In the suburban site T1 SIAlevels in PM2.5 were markedly reduced for nitrate due to themajor occurrence of coarser Ca and Na nitrate species. Thiswas also found by Fountoukis et al. (2007), who estimated30% of nitrate being in the coarse fraction during 21th–30thMarch 2006, similar to our 39% during the same period. It is



Figure 8

3N

5N

13N

15N

17N

21N

27N 29N

6N

0

2

4

6

8

10

12

14

SO

42-µ

g/m

3

0

20

40

60

80

100

120

140

V n

g/m

3

SO42- V

CENICA-PM 10

..

1D 3D 7D

9D 11D 13D 15D 17D

21D

25D

29D

31D 1D 2D 4D

19D

6D

0

2

4

6

8

10

12

01-03-06 06-03-06 11-03-06 16-03-06 21-03-06 26-03-06 31-03-06 05-04-06

NO

3-µ

g/m

3

0

5

10

15

20

25

30

35

Cru

sta

l µ

g/m

3

NO3- Crustal

..

3N

5N

13N

15N

17N

21N

27N 29N

6N

0

2

4

6

8

10

12

14

SO

42-µ

g/m

3

0

20

40

60

80

100

120

140

V n

g/m

3

SO42- V

CENICA-PM 10

..

1D 3D 7D

9D 11D 13D 15D 17D

21D

25D

29D

31D 1D 2D 4D

19D

6D

0

2

4

6

8

10

12

01-03-06 06-03-06 11-03-06 16-03-06 21-03-06 26-03-06 31-03-06 05-04-06

NO

3-µ

g/m

3

0

5

10

15

20

25

30

35

Cru

sta

l µ

g/m

3

NO3- Crustal

..

1D 3D 7D

9D 11D 13D 15D 17D

21D

25D

29D

31D 1D 2D 4D

19D

6D

0

2

4

6

8

10

12

01-03-06 06-03-06 11-03-06 16-03-06 21-03-06 26-03-06 31-03-06 05-04-06

NO

3-µ

g/m

3

0

5

10

15

20

25

30

35

Cru

sta

l µ

g/m

3

NO3- Crustal

..

Fig. 8. Mean 12 h levels of SO2−

4 , NO−

3 , V and crustal material recorded at CENICA during the MILAGRO campaign. N, maximum valuerecorded in the nighttime sample; D maximum value recorded in the daylight sample.

interesting to note that the above described fine PM episodes(overnight trapping of pollutants in Mexico City and duringthe morning hours) coincide with the peak levels of ammo-nium sulfate, V and Ni (tracers of fuel-oil combustion andregional/urban background PM climate). Thus, Figs. 8 and9 show the occurrence of these sulfate-vanadium episodes atCENICA, and Fig. 9 evidences a very good correlation be-tween the levels of the above tracers recorded at CENICAand T0. The levels of most crustal elements and Hg exhibita good correlation between the two urban sites. Figure 8also evidences that nitrate and crustal material episodes arerecorded during daytime and do not coincide with the aboveregional episodes, but are more related with local pollutionepisodes. This is also corroborated by Fig. 9, showing lowercorrelation between levels of nitrate, Ba and Zn (usually as-sociated with road traffic emissions and consequently higherat T0) measured at the two urban sites when compared withsulfate. This is also the case of As and Zr, with relativelyhigher levels being recorded at CENICA and with a low cor-relation with T0.

As expected themarine aerosols accounted for a very lowfraction of PM10 (<1.5% in most cases).

As previously reported for PM levels, with the excep-tion of carbonaceous aerosols and V, levels of major com-ponents andtrace elements in PM are usually much lower(one order of magnitude in some cases) during the MILA-GRO campaign than those reported from other large cities(Table 3 and Table 4). Levels in PM10 of As (5–7 ng/m3),Zn (100–500 ng/m3), Cd (1–3 ng/m3) and V (20–50 ng/m3)were relatively high when compared with most urban sites

from the US and Europe, or with values reported from Mex-ico city during 1996–1998 (Mugica et al., 2002) or duringMCMA-2003 campaign (Johnson et al., 2006). However,the limit and target values for Ni, Pb, Cd and As from theEU (1999/30/CE and 2004/107/CE) were met with the ex-ception of As (exceeding slightly the EU target value of 6ng/m3). Once again, it is essential to remember that the dataobtained during the MILAGRO campaign do not representannual mean values. The presence of metals in these sam-ples has also been observed in the SEM, with most of themcontaining Ba (irregular in shape and interpreted as derivedfrom traffic, probably brake linings), Fe (from condensedspherules to amorphous oxidized masses between 1–3µmin size) or Pb (usually 1µm or less in size, and commonlyspheroidal). Particles of other individual metals are muchrarer but include Cu, Zn, Mn, W, Sb, Ni (with V), and Mo,all of which are again very fine grained (usually sub-micron).

According to the above data, PM10 at the urban and in-dustrially influenced areas is mainly made up of OM+EC(32–46%), crustal material (25–27% urban, 15–35% indus-trial) and SIA (15–19% urban, 14–29% industrial). The un-accounted material (mostly water and heteroatoms in organicmatter) accounts for 3 to 29% of the PM10 mass, whereas thetrace elements account for 1 to 2% of the mass. At the ruralsites, the proportion of crustal material increases to 32–43%,whereas OM+EC diminishes down to 16–23% (Fig. 10). Theprevalence of carbonaceous aerosols in the PM compositiondetected in this study is consistent with the results by Chowet al. (2002) and Salcedo et al. (2006). In PM2.5 (Fig. 10) thecrustal material is reduced (but still present in relatively high


X. Querol et al.: PM speciation and sources in Mexico during MILAGRO campaign 121Figure 9

0

1

2

3

4

5

6

7

0 1 2 3 4 5 6 7

Al2O3 PM10 T0 (ng/m 3)

Al 2

O3 P

M1

0 C

EN

ICA

(n

g/m

3)

Al2O3

0

10

20

30

40

50

0 10 20 30 40 50

0

5

10

15

20

25

0 5 10 15 20 25

As PM10 T0 (ng/m 3)

As

PM

10 C

EN

ICA

(n

g/m

3)

As

Zr

Zr PM10 T0 (ng/m 3)

Zr

PM

10 C

EN

ICA

(n

g/m

3)

0

2

4

6

8

10

0 2 4 6 8 10

NO3-

NO3- PM10 T0 (ng/m 3)

NO

3- P

M1

0 C

EN

ICA

(n

g/m

3)

0

100

200

300

400

0 100 200 300 400

Ba PM10 T0 (ng/m 3)

Ba P

M10 C

EN

ICA

(n

g/m

3)

Ba

0

200

400

600

800

1000

0 200 400 600 800 1000

Zn PM10 T0 (ng/m 3)

Zn

PM

10 C

EN

ICA

(n

g/m

3)

Zn

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.0 0.2 0.4 0.6 0.8

Hg PM10 T0 (ng/m 3)

Hg

PM

10 C

EN

ICA

(n

g/m

3)

Hg

0

2

4

6

8

10

12

14

0 2 4 6 8 10 12 14

SO42- PM10 T0 (ng/m 3)

SO

42

- PM

10 C

EN

ICA

(n

g/m

3)

SO42-

0

50

100

150

0 50 100 150

V PM10 T0 (ng/m 3)

V P

M1

0 C

EN

ICA

(n

g/m

3)

V

Fig. 9. Correlation between levels of Al2O3, Zr, As, NO−

3 , Ba, Zn, Hg, V and SO2−

4 in PM10 at CENICA and T0.

proportions) in the urban area down to 15% and to 8–17%in the industrially influenced sites, but it is still present inaround 30% at the suburban site of T1 due to high local soilresuspension. Conversely, the proportions of carbonaceousaerosols increase up to 55% at the urban and industrial sitesand 30% at the suburban site.

As shown in Fig. 11 crustal elements are generally presentin the coarse fraction, especially at T1, where they are presentin the fraction>10µm in a significant proportion (PM10/TSPratio 04–0.6). The higher PM2.5/PM10 ratios were measuredat the urban and industrial sites (0.4 at T0 and Jasso) whencompared with the suburban (0.3) site, due to the major an-thropogenic origin of the mineral matter in the first locations(road dust, demolition, limestone quarry, cement plants, etc).The other elements, mostly with a major anthropogenic ori-gin, are mainly present in the fine fraction, with PM2.5/PM10and PM10/TSP ratios usually>0.6, but in many cases>0.8.However, in some cases, where soil resuspension is impor-tant (T1) some elements such as Ni may have an importantcoarse proportion (PM2.5/PM10 and PM10/TSP ratios closeto 0.3).

At the urban sites, levels of major and trace crustal compo-nents and nitrate were significantly higher during the daylightperiod than during the night (Fig. 8), by a factor day/nightranging from 1.0 to 1.9 (with the highest values at CENICA)for PM10, as a consequence of higher emissions (from roadtraffic and from resuspension). Conversely, as shown inFig. 8, levels of sulfate, chloride and most anthropogenictrace elements (Hg, V, Ni, Cu, Zn, As, Se, Mo, Cd, Sn, Baand Pb) were higher at night (by a factor of 0.8–1.0, 0.5–0.7and 0.4–0.9, respectively) probably due, as previously stated,to the reduction of the mixing layer depth. Carbonaceousaerosols increased as a mean also during the night at the ur-ban site T0 by a factor of 1.2 and at the suburban site T1 by afactor of 1.5. At the urban site CENICA day and night levelsof these components were very similar.

According to the transport and synoptic scenarios de-scribed by Fast et al. (2007) for the MILAGRO campaign, at-mospheric transport form Mexico City over the suburban siteT1 occurred on 10th, 18th, 19th, 20th, 22nd, and 24th March2006. From these days sampling took place only during 9th–10th and 19th–20th. As it may be observed in Fig. 12, theatmospheric transport from Mexico City is traced at T1 bythe increase of nitrate and potassium in PM2.5, whereas the



0%

20%

40%

60%

80%

100%

CENICA T0 UNAM Jasso PXT T2 T1

OM EC OM+EC Crustal SO42- NO3- NH4+ Trace elements UnaccountedSO42- NO3

- NH4+

0%

20%

40%

60%

80%

100%

T0 Jasso PXT T1

OM EC Crustal SO42- NO3- NH4+ Trace elements UnaccountedSO42- NO3

- NH4+

Fig. 10. Mean PM10 (top) and PM2.5 (bottom) composition obtained at urban (dark grey), industrial-rural (clear grey) and suburban andrural sites. Separate OM and EC data were unavailable for T2 site, and therefore their sum (OM+EC) is plotted for this site only.

transport from Tula industrial estate is traced by the increaseof sulfate levels on 21st March 2006.

4.3 Statistical analysis

Comparing the mean levels oftrace elements measured atthe different study sites in the present work and with the dis-tribution of major PM components, the following source ori-gins may be identified:

– Crustal origin (soil and/or urban dust): Li, Be, P, Sc,Ti, Mn, Co, Ga, Ge, Rb, Sr, Y, Zr, Nb, Cs, Rare EarthElements, Hf, U, Th.

– Urban (mostly road traffic): Cr, Mn, Cu, Zn, As, Se, Cd,Sn, Sb, Ba and Pb, typically associated with road traf-fic emissions (engine emissions, abrasion of tyres andbrake pads).

– Industrial around Tula: Pb, Cr and V (other elementssuch as Cd, As, Ni were not analyzed here).

– Industrial and regional (petrochemical estate, fuel oil,petcoke combustion): V, Ni and sulfate.

Particulate Hg is not higher at the urban sites than at the sub-urban site T1, consequently, as previously reported, it has anexternal (regional) origin to the city. However, local sources,such as a hazardous waste incinerator and a petrochemicalplant may also contribute to increase Hg in a much minorproportion (Gonzalez et al., 2007).

These findings were confirmed by performing a statisti-cal analysis of the different datasets, by means of Princi-pal Component Analysis (PCA) with the software STATIS-TICA v.4.2. This kind of factor analysis requires a mini-mal internal variability, and therefore it was only applied tothe datasets containing>25 samples (CENICA, T0 and T1).However, the number of samples (25, 35 and 38 at T1, T0and CENICA, respectively) was lower than the minimumnumber of cases required according to Henry et al. (1984)(n>30+(V+3)/2, where n is the number of samples and V isthe number of variables). Consequently, in order to confirm



Table 4. Mean TSP, PM10 and PM2.5 levels and major components (µg/m3) recorded at the study sites during the MILAGRO campaign.

CENICA T0 T1 T2 UNAM Jasso PXT

PM10 PM10 PM2.5 PST PM10 PM2.5 PM10 PM10 PM10 PM2.5 PM10 PM2.5n 38 36 27 23 24 25 8 7 28 28 27 27PM 53 50 40 193 82 33 34 54 75 31 37 26OC 10.8 12.9 12.4 8.0 5.0 3.7 ND 8.0 12.9 8.0 7.1 6.8EC 3.6 2.7 2.1 5.6 5.3 4.0 ND 4.7 6.0 3.6 2.2 2.2CO2−

3 2.0 2.3 0.9 9.2 4.2 1.1 1.3 2.3 13.8 2.5 1.4 0.5SiO2 5.1 4.9 2.4 34.6 15.7 4.0 4.7 5.6 2.4 0.6 1.4 0.9Al2O3 2.5 2.4 1.2 17.3 7.9 2.0 2.3 2.8 0.6 0.1 0.4 0.1Ca 1.3 1.5 0.6 6.2 2.8 0.7 0.9 1.5 9.2 1.7 0.9 0.3K 0.7 0.7 0.3 1.9 1.1 0.6 0.5 0.6 0.2 0.1 0.1 0.2Na 0.5 0.5 0.3 1.6 0.8 0.3 0.2 0.4 0.2 0.1 0.1 0.1Mg 0.5 0.4 0.1 1.8 0.9 0.3 0.3 0.4 0.3 0.1 0.1 0.04Fe 0.9 1.1 0.4 4.2 2.0 0.6 0.6 1.0 0.6 0.1 0.6 0.2SO2−

4 4.9 5.3 3.7 6.3 4.8 3.9 4.4 4.2 6.4 3.7 6.4 6.4NO−

3 3.5 3.6 2.8 4.2 2.8 1.1 1.2 1.8 1.8 0.6 1.6 0.8Cl− 0.4 0.4 0.7 0.5 0.3 0.2 0.1 0.1 0.1 0.2 0.1 0.1NH+

4 1.6 1.5 1.0 1.2 1.1 1.1 1.2 1.5 2.0 1.6 2.4 2.5

the validity of the results PCA was applied separately to thePM10 datasets from T0, CENICA and T1, and it was also ap-plied to a larger dataset made up of the combination of theCENICA and T0 samples, with similar results. This con-firmed the robustness of the solution.

The results for CENICA were very similar to those for T0,and thus are not included in the description below. At theurban locations (CENICA and T0), four main factors wereidentified (representing the main source categories):

Factor 1:the main tracers of this factor were Al2O3, Ti, Mg, Fe, Ca,P, K, Na, Mn, Th, Ce, Zr, suggesting the clearly mineral ori-gin of this source. This factor probably represents mineraldust resuspension in the city, both road/city and soil dust.The daily evolution of the contribution to the PM mass ofthis factor showed a marked decrease after 23rd March 2006,coinciding with the washout of road and city dust by precip-itation.

Factor 2:this factor was characterized by SO2−

4 , V, Ni, NH+

4 and Hg,and it was interpreted as regional-scale transport. The contri-bution from SO2−

4 and NH+

4 suggests long-range transport,and once again the correlation between these species and Hgis observed and it confirms the regional origin of Hg. Thefact that Ni and V are also found in this factor shows thatthe impact from fuel-oil combustion emissions such as thoseregistered in the Tula industry estate may be detected in Mex-ico City under specific air mass transport scenarios (regional-scale transport).

Factor 3:Sr, Cu, Ba, Pb, Cr, C, Zn and Cd were the main tracers of thissource, which was interpreted as road traffic. SEM-EDX ob-servations confirmed this association of metals. This sourcewas the only one presenting slight differences between T0and CENICA, probably due to the larger traffic influencein T0. Whereas the factor loadings of major and primarytraffic tracers were higher at T0 (C, Cu, Ba, Pb), it was

the secondary or minor traffic tracers that were highest atCENICA (NO−

3 , Rb, Cd). This confirms that even thoughtraffic emissions are dominant at both urban sites, their im-pact is more direct in the case of T0 than of CENICA.

Factor 4:the last factor was characterized by As, Zr, Th, Ce, and Cr,and the nature of this source is yet unclear. This combinationof elements could suggest the resuspension of volcanic dustdeposited on the city roads, or the influence of an additionalsource affecting both urban sites.

In the case of the T1 suburban site, five factors were iden-tified:

Factor 1:with Fe, Ti, Mn, Al2O3, Ce, Mg, Th, K, Rb, Sr, Ba, Na, Zr,Cr, Ca and P as main tracers, the nature of this source wasinterpreted as mineral, although some differences may be ob-served with the mineral source in T0 and CENICA. While thechemical profile in T1 suggests the prevalence of clay miner-als, in T0 and CENICA a mixture is observed between clayminerals, carbonates and K-feldspars among others (higherfactor loadings for Ca, P, K) originating from constructionand demolition activities in the city.

Factor 2:this factor was characterized by Pb, Cd, Cu and Zn and itrepresents an anthropogenic source of PM, probably linkedto industrial emissions which could be related to small smelt-ing plants located to the northeast of Mexico City (R. Mof-fett, personal communication).

Factor 3:with NH+

4 and SO2−

4 as its main tracers, this source clearlyrepresents regional-scale transport. Hg was also found to cor-relate with this source, even though the highest factor loadingfor this element was found in Factor 4.

Factor 4:Sb, As, C, NO−3 and Cd were the main tracers of this source,which coincided largely with the traffic source at the ur-ban sites. Thus, this source could be interpreted as the



Table 5. Mean levels of trace components in TSP, PM10 and PM2.5 (ng/m3) recorded at the study sites during the MILAGRO campaign.NA: not available.

CENICA T0 T1 T2 UNAM Jasso PXT

PM10 PM10 PM2.5 TSP PM10 PM2.5 PM10 PM10 PM10 PM2.5 PM10 PM2.5n 38 36 27 23 24 25 8 7 28 28 27 27Hg 0.14 0.22 0.13 0.40 0.28 0.18 0.04 0.11 NA NA NA NALi 0.4 0.5 0.4 2.2 1.0 0.2 0.1 0.5 NA NA NA NABe 0.05 0.05 0.04 0.57 0.27 0.12 0.05 0.05 NA NA NA NAP 100 120 53 304 166 48 72 106 NA NA NA NASc 0.6 0.6 0.5 1.6 1.0 0.5 0.2 1.1 NA NA NA NATi 109 81 36 440 205 57 65 85 NA NA NA NAV 19 25 13 27 17 10 9 16 50 30 21 18Cr 0.7 4.4 2.1 4.9 1.5 0.9 0.1 1.9 20 15 192 102Mn 20 32 16 83 41 13 15 23 NA NA NA NACo 0.5 0.8 0.5 1.6 0.9 0.2 0.3 0.5 NA NA NA NANi 3 5 3 9 5 2 1 3 NA NA NA NACu 78 110 90 103 33 27 13 140 13 11 10 5Zn 103 482 244 187 165 97 8 98 21 11 11 5Ga 0.4 0.4 0.3 1.9 1.2 0.4 0.4 0.1 NA NA NA NAGe <0.1 0.2 0.2 0.3 0.1 0.1 <0.1 <0.1 NA NA NA NAAs 6 6 6 4 3 2 1 7 NA NA NA NASe 7 6 6 3 3 3 1 7 NA NA NA NARb 0.9 0.9 0.5 2.5 1.8 1.0 0.6 0.7 NA NA NA NASr 9 16 5 38 18 5 5 9 NA NA NA NAY 0.3 0.2 0.1 1.8 0.9 0.2 0.1 0.2 NA NA NA NAZr 20 18 15 45 29 22 6 23 NA NA NA NANb 0.4 0.4 0.3 1.9 1.1 0.4 0.3 0.5 NA NA NA NAMo 1 2 2 8 5 3 0.1 0.1 NA NA NA NACd 2 3 3 1 1 1 0.3 2 NA NA NA NASn 23 32 29 12 14 9 4 7 NA NA NA NASb 15 16 15 10 10 8 3 8 NA NA NA NACs 0.1 0.1 0.1 0.3 0.2 0.1 <0.1 0.2 NA NA NA NABa 30 92 22 82 43 12 3 27 NA NA NA NALa 1.3 0.8 0.7 0.4 1.2 0.5 0.7 0.5 NA NA NA NACe 1.6 1.5 1.2 6.9 4.1 1.1 1.0 1.5 NA NA NA NAHf 0.7 0.6 0.5 1.4 0.9 0.7 0.2 0.8 NA NA NA NATl 0.6 0.4 0.3 0.3 0.5 0.2 <0.1 0.3 NA NA NA NAPb 34 111 62 35 31 23 6 26 23 20 253 199Bi 0.2 0.4 0.3 0.3 0.1 0.1 0.2 0.2 NA NA NA NATh 0.1 0.1 0.1 1.1 0.5 0.1 0.1 0.2 NA NA NA NAU 0.3 0.3 0.2 0.5 0.4 0.4 0.2 0.4 NA NA NA NA

contribution from local traffic but also as the transport to-wards T1 of the emissions from Mexico City. This factor hasalso high loadings for K and Hg.

Factor 5:V and Ni were the main tracers of this source, which wasinterpreted as fuel-oil combustion probably from the petro-chemical estate in Tula. It is interesting to observe that at T1the regional source does not include the contributions fromV and Ni, which were detected jointly at the urban sites.This indicates that the impact from the emissions from thepetrochemical estate is registered at T1 under specific airmass transport scenarios, which do not always coincide withregional-scale transport as in the case of the urban sites.

It is interesting to note that biomass burning emissionswere not identified as a single source in any of the factoranalyses performed. Instead, tracers of these emissions (K)were correlated with the mineral and regional-fuel oil com-bustion sources. The non-identification of this source maybe related to its probably low contribution to the PM10 masscompared to other sources during the study period. Also thelimited number of samples available (25–38 samples/site),

the short duration of the campaign, and the fact that spe-cific tracers (OC, WSOC, water-soluble K) were not avail-able for all samples may account for the non identification ofthis source. However, the correlation between OC+EC andK levels described above (Sect. 4.2) evidences the influenceof this source on PM levels and composition. Potassium inPM2.5 was also detected as tracer of the Mexico’s plume ata rural site in this study. Thus the contribution of biomassburning to PM10 and PM2.5 was detected but in much minorproportions than reported by other authors by modeling. Fur-thermore, for finer aerosols (such as PM1) this source mayprobably contribute in a larger extend due to the reduction ofthe mineral matter load (containing also K).

5 Conclusions

The levels of PM during the MILAGRO campaign (1st to31st March 2006) ranged from 50 to 56µg PM10/m3, 24 to40µg PM2.5/m3 and 19 to 33µg PM1/m3 in Mexico City.Although these levels may be considered as relatively high



0.0

0.2

0.4

0.6

0.8

1.0

Li

Be P

Sc Ti

Cr

Mn

Co

Ga

Rb

Sr Y Zr

Nb

Ba

Ce Pr

Nd

Hf

Ta

Th V Ni

Hg

Cu

Zn

As

Se

Mo

Cd

Sn

Sb

Cs

Pb Bi U

T0 PM2.5/PM10 T1 PM2.5/PM10 PXT PM2.5/PM10 Jasso PM2.5/PM10

M ineral elements Other elements

PM2.5

/PM10

PM2.5

/PM10

PM2.5

/PM10

PM2.5

/PM10

PM

2.5/P

M10 r

atio

0.0

0.2

0.4

0.6

0.8

1.0

Li

Be P

Sc Ti

Cr

Mn

Co

Ga

Rb

Sr Y Zr

Nb

Ba

Ce Pr

Nd

Hf

Ta

Th V Ni

Hg

Cu

Zn

As

Se

Mo

Cd

Sn

Sb

Cs

Pb Bi U

T1 PM2.5/PM10 T1 PM10/TSP

M ineral elements Other elements

PM2.5

/PM10

PM10

/TSP

PM

2.5/P

M10 a

nd

PM

10/T

SP

ra

tio

s

Fig. 11. PM2.5/PM10 ratios of trace elements levels at T0, T1, Jasso and PXT. PM10/TSP levels at T1.

W windsN&E winds W&SW winds variable SW winds

0

1

2

3

4

01-03-06 06-03-06 11-03-06 16-03-06 21-03-06 26-03-06 31-03-06

NO

3-

an

dK

µg

/m3

0

2

4

6

8

10

12

SO

42

-µg

/m3

NO3- K SO4

2-

T1-PM 2.5

..

Transport from Tula industrial estate

Transport from México city

Fig. 12.12 h levels (daylight and night) of nitrate, potassium and sulfate in PM2.5 measured at the T1 suburban site on 1st, 3rd, 5th, 7th, 9th,11th, 13th, 16th, 17th, 19th, 21st, 23rd, 25th, 27th and 29th March 2006. Note that nitrate and potassium trace the transport from MexicoCity, whereas sulfate is mainly tracing the transport from Tula industrial estate. The transport and synoptic scenarios described by Fast etal. (2007) for the MILAGRO campaign are also marked in the plot.

when compared with US or European air quality standards,they are markedly lower than PM levels measured at Asianmega-cities.

Daily variations of PM levels were recorded in parallel at

the urban sites within the city, despite the distance betweenthe sites (approx. 20 km). This suggests that the variabil-ity of PM levels and composition in Mexico City are de-termined by atmospheric dynamics, more specifically by the



mixing layer height, rather than by emission sources. StrongPM peak episodes originated by the decreasing mixing layerheight were detected on a number of days between 04:00 and06:00 h.

Important day-night variations were also observed: thelevels of crustal components and NO−

3 presented higher lev-els during daytime (coinciding with Chow et al., 2002 andSalcedo et al., 2006), while SO2−

4 concentrations and mostof anthropogenic trace metals presented higher levels duringnight hours.

The analysis of air mass transport scenarios demonstratedthe impact of the emissions from Mexico City on the lev-els and chemical composition of PM on surrounding areas.On days when winds originated from the north/north-west,the impact of the Mexico City plume was clearly detected atthe rural background site in Tenango. Conversely, under theinfluence of winds from the south/south-east, the influenceof the Mexico City plume was detected at the T1 site, withnitrate and potassium in PM2.5 being identified as major trac-ers. Finally, westerly circulations evidenced the influence ofthe emissions from Tula (petrochemical estate) on the chem-ical composition of PM at T1.

The statistical analysis of the PM10 data from T0,CENICA and T1 by PCA resulted in four sources at theurban sites (CENICA and T0), and five sources at thesuburban site (T1). At the urban locations the main PMsources were mineral dust (both road/city and soil dust),regional-scale transport (including Hg), traffic emissionsand an unidentified As source. Traffic emissions werecharacterized by primary tracers at T0, whereas secondarytracers were highest at CENICA, confirming that theimpact of traffic emissions is more direct in T0 than inCENICA. At T1 the main PM sources were mineral dust,industrial emissions, regional-scale transport, local traffic(also interpreted as the transport towards T1 of the emis-sions from Mexico City), and oil combustion originatingprobably from the petrochemical estate in Tula. The inter-pretation of the traffic source at all sites was confirmed bythe observation of Ba, Fe, Cu and Pb particles by SEM-EDX.

Supplementary material: http://www.atmos-chem-phys.net/8/111/2008/acp-8-111-2008-supplement.zip

Acknowledgements. This study was supported by the AccionComplementaria from the Ministry of Education and Science ofSpain CGL2005-23745-E. We also thank the support from theSpanish Ministry of the Environment. We would like also to thankW. Eichinger from the University of Iowa for making available themeteorological data at T0 during the campaign.

Edited by: L. Molina

References

Aldape, F., Flores M. J., Flores A. J., Retama-Hernandez, A., andRivera-Hernandez, O: Elemental composition and source identi-fication of PM2.5 particles collected in downtown Mexico City,International Journal PIXE, 15, 263-270, 2005.

Balachandran, S., Meena, B. R., and Khillare, P. S.: Particle sizedistribution and its elemental composition in the ambient air ofDelhi, Environ. Int., 26, 49–54, 2000.

Bogo, H., Otero, M., Castro, P., Ozafran, M. J., Kreiner, A., Calvo,E. J., Negri, R. M.: Study of atmospheric particulate matter inBuenos Aires city, Atmos. Environ., 37, 1135–1147, 2003.

CAFE: Clean Air for Europe: European Commission, Second Po-sition Paper on Particulate Matter – CAFE Working Group onParticulate Matter, D.G. Environment, April 2004, 1–234, 2004.

Chan, C. Y., Xu, X. D., Li, Y. S., Wong, K. H., Ding, G. A., Chan,L. Y., and Cheng, X. H.: Characteristics of vertical profiles andsources of PM2.5, PM10 and carbonaceous species in Beijing,Atmos. Environ., 39, 5113–5124, 2005.

Chow, J. C., Watson, J. C., Edgerton S. A., and Vega, E.: Chemicalcomposition of PM2.5 and PM10 in Mexico City during winter1997, Sci. Total Environ., 287, 177–201, 2002.

Clean Air Initiatives for Asian cities: Air Quality in Asiancities, 1–6, available at:http://www.cleanairnet.org/caiasia/1412/article-59689.html, 2006.

Dockery D., Pope C., Xu, X., Spengler, J., Ware, J., Fay, M., Ferris,B., and Speizer, F.: An association between air pollution andmortality in six US cities, New Engl. J. Med., 329, 1753–1759,1993.

Doran, J. C., Arnott, W. P., Barnard, J. C., Cary, R., Coulter, R.,Fast, J. D., Kassianov, E. I., Kleinman, L., Laulainen, N. S., Mar-tin, T., Paredes-Miranda, G., Pekour, M. S., Shaw, W. J., Smith,D. F., Springston, S. R., and Yu, X.-Y.: The T1-T2 study: evo-lution of aerosol properties downwind of Mexico City, Atmos.Chem. Phys., 7, 1585–1598, 2007,http://www.atmos-chem-phys.net/7/1585/2007/.

Fast, J. D. and Zhong, S.: Meteorological factors associated withinhomogeneous ozone concentrations within the Mexico Citybasin, J. Geophys. Res., 103, 18 927–18 946, 1998.

Fast, J. D., de Foy, B., Acevedo Rosas, F., Caetano, E., Carmichael,G., Emmons, L., McKenna, D., Mena, M., Skamarock, W.,Tie, X., Coulter, R. L., Barnard, J. C., Wiedinmyer, C., andMadronich, S.: A meteorological overview of the MILAGROfield campaigns, Atmos. Chem. Phys., 7, 2233–2257, 2007,http://www.atmos-chem-phys.net/7/2233/2007/.

Fountoukis, C., Nenes, A., Sullivan, A., Weber, R., VanReken, T.,Fischer, M., Matıas, E., Moya, M., Farmer, D., and Cohen, R. C.:Thermodynamic characterization of Mexico City aerosol duringMILAGRO 2006, Atmos. Chem. Phys. Discuss., 7, 9203–9233,2007,http://www.atmos-chem-phys-discuss.net/7/9203/2007/.

Fung, K, Chow, J, and Watson, J.: Evaluation of OC/EC Specitionby Thermal Manganese Dioxide Oxidation and the IMPROVEMethod, Air Waste Manage., 52, 1333–1341, 2002.

Gonzalez, R., Marquez, C., Bernabe, R., Snyder, D. C., Rutter, A.P., Minguillon, M. C., Perez, N., Pey, J., Alverdin, L., Reyes, E.,Miranda, L., Miranda, J., Moreno, T., Blanco, S., De la Rosa, A.,Solorzano, G., Alastuey, A., Schauer, J., Querol, X., de Foy, B.,Molina, L. T., and Cardenas, B.: A43A-20 Measurements of To-tal Gaseous and Particulate Mercury in Mexico City During the


http://www.atmos-chem-phys.net/8/111/2008/acp-8-111-2008-supplement.zip

http://www.atmos-chem-phys.net/8/111/2008/acp-8-111-2008-supplement.zip



http://www.atmos-chem-phys.net/7/1585/2007/


http://www.atmos-chem-phys-discuss.net/7/9203/2007/


MCMA-2006/MILAGRO Campaign. Eos Trans. AGU, 88(23),Jt. Assem. Suppl., Abstract A43A-20, 2007.

Hea, K., Yang, F., Ma, Y., Zhang, Q., Yao, X., Chan, Ch.K., Cadle,S., Chan, T., and Mulawa, P.: The characteristics of PM2.5 inBeijing, China, Atmos. Environ., 35, 4959–4970, 2001.

Henry, R. C., Lewis, C. W., Hopke, P. K., and Williamson, H. J.:Review of receptor model fundamentals, Atmos. Environ., 18,1507–1515, 1984.

Ho, K. F., Lee, S. C., Chan, C. K., Yu, J.C., Chow, J. C., and Yao,X.H.: Characterization of chemical species in PM2.5 and PM10aerosols in Hong Kong, Atmos. Environ., 37, 31–39, 2003.

Johnson, K. S., de Foy, B., Zuberi, B., Molina, L. T., Molina, M.J., Xie, Y., Laskin, A., and Shutthanandan, V.: Aerosol composi-tion and source apportionment in the Mexico City MetropolitanArea with PIXE/PESA/STIM and multivariate analysis, Atmos.Chem. Phys., 6, 4591–4600, 2006,http://www.atmos-chem-phys.net/6/4591/2006/.

Karar, K. and Gupta, A. K.: Source apportionment of PM10 at res-idential and industrial sites of an urban region of Kolkata, India,Atmos. Res., 84, 30–41, 2007.

Kim, K.-H. and Kim, M.-Y.: The effects of Asian Dust on partic-ulate matter fractionation in Seoul, Korea during spring 2001,Chemosphere, 51, 707–721, 2003.

Kunzli, N., Kaiser, R., Medina, S., Studnicka, M., Chanel, O.,Filliger, P., Herry, M., Horak, F. Jr., Puybonnieux-Texier, V.,Quenel, P., Schneider, J., Seethaler, R., Vergnaud, J.-C., andSommer, H.: Public health impact of outdoor and traffic-relatedair pollution: a tri-national European assessment, Lancet, 356,795–801, 2000.

Kumar, A. V., Patil, R. S., and Nambi, K. S. V.: Source apportion-ment of suspended particulate matter at two traffic junctions inMumbai, India, Atmos. Environ., 35, 4245–4251, 2001.

Miranda, J., Andrade, E., Lopez-Suarez, A., Ledesma, R., Cahill, T.A., and Wakabayashi, P. H.: A receptor model for atmosphericaerosols from a southwestern site in Mexico City, Atmos. Envi-ron., 30, 3471–3479, 1996.

Miranda, J., Barrera, V. A., Espinosa, A. A., Galindo, O. S., andMeinguer, J.: PIXE analysis of atmospheric aerosols in MexicoCity, X-Ray Spectrom., 34, 391–401, 2005.

Moffett, R. C., de Foy, B., Molina, L. T., Molina, M. J., and Prather,K. A.: Measurement of ambient aerosols in northern Mexico Cityby single particle mass spectrometry, Atmos. Chem. Phys. Dis-cuss., 7, 6413–6457, 2007,http://www.atmos-chem-phys-discuss.net/7/6413/2007/.

Molina, L. T., Kolb, C. E., de Foy, B., Lamb, B. K., Brune, W.H., Jimenez, J. L., Ramos-Villegas, R., Sarmiento, J., Paramo-Figueroa, V. H., Cardenas, B., Gutierrez-Avedoy, B., and Molina,M. J.: Air quality in North America’s most populous city –overview of the MCMA-2003 campaign, Atmos. Chem. Phys.,7, 2447–2473, 2007,http://www.atmos-chem-phys.net/7/2447/2007/.

Monkkonen, P., Uma, R., Srinivasan, D., Koponen, I. K., Lehtinen,K. E. J., Hameri, K., Suresh, R., Sharma, V. P., and Kulmala,M.: Relationship and variations of aerosol number and PM10mass concentrations in a highly polluted urban environment—New Delhi, India, Atmos. Environ., 38, 425–433, 2004.

Mugica, V., Maubert, M., Torres, M., Munoz, J., and Rico, E.: Tem-poral and Spatial Variations of Metal Content in TSP and PM10in Mexico City during 1996–1998, J. Aerosol Sci., 33(1), 91–

102, 2002.Querol, X., Alastuey, A., Rodrıguez, S., Plana, F., Ruiz, C. R., Cots,

N., Massague, G., and Puig, O.: PM10 and PM2.5 source appor-tionment in the Barcelona Metropolitan Area, Catalonia, Spain,Atmos. Environ., 35/36, 6407–6419, 2001.

Rojas-Bracho, L., Suh, H. H., Oyola, P., and Koutrakis, P.: Mea-surements of children’s exposures to particles and nitrogen diox-ide in Santiago, Chile, The Science of The Total Environment,287, 3, 249-264, 2002.

Salcedo, D., Onasch, T. B., Dzepina, K., Canagaratna, M. R.,Zhang, Q., Huffman, J. A., DeCarlo, P. F., Jayne, J. T., Mor-timer, P., Worsnop, D. R., Kolb, C. E., Johnson, K. S., Zuberi,B., Marr, L. C., Volkamer, R., Molina, L. T., Molina, M. J., Car-denas, B., BernabE, R. M., Marquez, C., Gaffney, J. S., Marley,N. A., Laskin, A., Shutthanandan, V., Xie, Y., Brune, W., Lesher,R., Shirley, T., and Jimenez, J. L.: Characterization of ambientaerosols in Mexico City during the MCMA-2003 campaign withAerosol Mass Spectrometry: results from the CENICA Super-site, Atmos. Chem. Phys., 6, 925–946, 2006,http://www.atmos-chem-phys.net/6/925/2006/.

Stone, E. A., Snyder, D. C., Sheesley, R. J., Sullivan, A. P., Weber,R. J., and Schauer J. J.: Source apportionment of fine organicaerosol in Mexico City during the MILAGRO Experiment 2006,Atmos. Chem. Phys. Discuss., 7, 9635–9661, 2007,http://www.atmos-chem-phys-discuss.net/7/9635/2007/.

Sun, Y., Zhuang, G., Wang, Y., Han, L., Guo, J., Dan, M., Zhang,W., Wang, Z., and Hao, Z.: The air-borne particulate pollutionin Beijing concentration, composition, distribution and sources,Atmos. Environ., 38, 5991–6004, 2004.

Turpin, B. J. and Lim, H.J.: Species Contributions to PM2.5 MassConcentrations: Revisiting Common Assumptions for Estimat-ing Organic Mass, Aerosol Sci. Tech., 35, 602–610, 2001.

Vega, E., Mugica, V., Reyes, R., Sanchez, E., Chow, J., and Watson,J.: Chemical Composition of Fugitive Dust Emitters in MexicoCity, Atmos. Environ., 35, 4033–4039, 2001.

Vega, E., Reyes, E., Ruiz, H., Garcia, J., Sanchez, G., Martinez-Villa, G., Gonzalez, U.: Analysis of PM2.5 and PM10 in the at-mosphere of Mexico City during 2000–2002, J. Air Waste Man-age., 54(7), 786–798, 2004.

Vega, E., Ruiz, H., Martinez-Villa, G., Sanchez, G., Sosa, G.,Gonzalez-Avalos, E., Reyes, E., and Garcia, J.: Fine and CoarseParticulate Matter Chemical Characterization in a Heavily Indus-trialized City in Central Mexico during Winter 2003, J. Air WasteManage., 57, 620–633, 2007.

Volkamer, R., Jimenez, J. L., SanMartini, F., Dzepina, K., Zhang,Q., Salcedo, D., Molina, L. T., Worsnop, D. R., and Molina, M.J.:, Secondary organic aerosol formation from anthropogenic airpollution: Rapid and higher than expected, Geophys. Res. Lett.,33, L17811, doi:10.1029/2006GL026899, 2004.

Wang, G., Wang, H., Yu, Y., Gao, Sh., Feng, J., Gao, S., and Wang,L.: Chemical characterization of water-soluble components ofPM10 and PM2.5 atmospheric aerosols in five locations of Nan-jing, China, Atmos. Environ., 37, 2893–2902, 2003.

WHO: Health aspects of air pollution with particulate matter,ozone and nitrogen dioxide, World Health Organization, ReportEUR/03/5042688, 1–98, 2003.

WHO: World Health Organization air quality guidelines global up-date 2005, 1–25, 2005.

WHO/UNEP: Urban Air Pollution in Megacities of the World,








Blackwell, Oxford, 1992.Yokelson, R., Urbanski, S., Atlas, E., Toohey, D., Alvarado, E.,

Crounse, J., Wennberg, P., Fisher, M., Wold, C., Campos, T.,Adachi, K., Buseck, P. R., and Hao, W. M.: Emissions fromforest fires near Mexico City, Atmos. Chem. Phys. Discuss., 7,6687-6718, 2007,http://www.atmos-chem-phys-discuss.net/7/6687/2007/.

Zhang, Q., Jimenez, J. L. Canagaratna, M. R., Allan, J. D., Coe,H. Ulbrich, I., Alfarra, M. R., Takami, A., Middlebrook, A.M., Sun, Y.L., Dzepina, K., Dunlea, E., Docherty, K., De-Carlo, P. F., Salcedo, D., Onasch, T., Jayne, J. T., Miyoshi,T., Shimono, A., Hatakeyama, S., Takegawa, N., Kondo, Y.,Schneider, J., Drewnick, F., Borrmann, S., Weimer, S., Demer-jian, K., Williams, P., Bower, K., Bahreini, R., Cottrell, L.,Griffin, R. J., Rautiainen, J., Sun, J. Y., Zhang, Y. M., andWorsnop, D. R.: Ubiquity and dominance of oxygenated speciesin organic aerosols in anthropogenically-influenced NorthernHemisphere midlatitudes, Geophys. Res. Lett., 34, L13801,doi:10.1029/2007GL029979, 2007.

Zheng, J., Tan, M., Shibata, Y., Tanaka, A., Li, Y., Zhang, G.,Zhang, Y., and Shan, Z.: Characteristics of lead isotope ratiosand elemental concentrations in PM10 fraction of airborne par-ticulate matter in Shanghai after the phase-out of leaded gasoline,Atmos. Environ., 38, 1191–1200, 2004.



pm speciation and sources in mexico during the milagro-2006

Documents